Thread: pg document is not recorded:maintenance_work_mem can not large than 2GB in windows

pg document is not  recorded:maintenance_work_mem can not large than 2GB in windows。

https://www.postgresql.org/docs/14/runtime-config-resource.html


I suggest that:modify maintenance_work_mem parameter description in the pg documents:

add maintenance_work_mem parameter size limit in windows platform


if you set maintenance_work_mem large that 2GB in windows platform

postgresql service can not start。 

in the windows event viewer ,thare is the following error:

"ERROR: 3072000 is outside the valid range for parameter "maintenance_work_mem" (64 .. 2097151)


and the work_mem parameter also has similar problems as above。

https://www.postgresql.org/message-id/17895.1315869622@sss.pgh.pa.us


so, It is best to add this description to the document about work_mem and maintenance_work_mem parameter


On Sun, Jul 24, 2022, at 1:01 AM, yanliang lei wrote:

pg document is not  recorded:maintenance_work_mem can not large than 2GB in windows。

https://www.postgresql.org/docs/14/runtime-config-resource.html


I suggest that:modify maintenance_work_mem parameter description in the pg documents:

add maintenance_work_mem parameter size limit in windows platform


if you set maintenance_work_mem large that 2GB in windows platform

postgresql service can not start。 

in the windows event viewer ,thare is the following error:

"ERROR: 3072000 is outside the valid range for parameter "maintenance_work_mem" (64 .. 2097151)


and the work_mem parameter also has similar problems as above。

https://www.postgresql.org/message-id/17895.1315869622@sss.pgh.pa.us

I don't think it is a good practice to document operating system limitations
because these limitations may be resolved in later versions of the operating
system. In this case, an OS choice (int and long are 32-bit values on 64-bit
Windows) imposes a Postgres limitation. The documentation does not mention an
upper limit and it also provides a suitable error message. Both are sufficient
to provide proper functioning for Postgres. IMO that's a change that someone
might be excited to address in the future.




--
Euler Taveira